Optical transmission systems generally use on/off switching or keying to generate a series of pulses. The presence of light from a light source commonly indicates a first state (e.g., a “1”) and the absence of light indicates a second state (e.g., a “0”. As a light source, the semiconductor laser has distinct advantages over light-emitting diodes or other sources. The semiconductor laser can support higher data rates and can reach longer distances when the emitted light is transmitted via a glass fiber. Due to the physics of the turn on transition, namely wavelength shift and wavelength distortion that result from modulating the laser through the lasing threshold, direct modulation is problematic. One way of minimizing the problems introduced by modulating a laser through the lasing threshold is to bias the laser source such that the OFF state is not completely off. While biasing the laser source avoids the performance degradation (wavelength shift, waveform distortion) that results from modulating the laser through the lasing threshold, introducing the bias creates difficulties in generating precise calibration signals to measure performance parameters of the optical transmitter.
A key indicator of the performance of an optical-fiber-based communications system is the extinction ratio. The extinction ratio describes the efficiency with which the transmitted optical power is modulated over the fiber-optic transport. The extinction ratio is the relationship of the power used in transmitting a logic level “1” (or P1) to the power used in transmitting a logic level “0” (or P0). The extinction ratio can be defined as a linear ratio, P1/P0; as a power measurement, 10× log(P1/P0); or as a percentage, (P0/P1)×100.
A measurement system for determining the extinction ratio of an optical transmitter under test typically consists of a digital storage oscilloscope or wideband digital-sampling oscilloscope and an optical to electrical converter. In practice, an accurate determination of extinction ratio is difficult to make. The response characteristics of the oscilloscope or other test equipment used when measuring signal components received from an optical to electrical signal converter are generally confirmed with a high-precision calibration signal. These prior art calibration methods supply a calibration signal with known DC (direct current) and AC (alternating current) values or a known relation between the DC and AC components of the calibration signal to the common input of DC and AC measuring devices. Such a calibration signal can be defined by frequency or the period, the peak-to-peak amplitude, the DC offset and an average value. A DC offset is that portion or component of the signal that is shifted or offset in amplitude from a reference value and does not vary with time. A DC measuring device provides a representation of the average amplitude of the signal. An AC measuring device may provide one or both of the frequency and a scale representation of the peak-to-peak amplitude of the signal.
FIG. 1 illustrates an example waveform generator 10 that generates a calibration signal with known DC and AC values or a known relationship between the DC and AC components of the generated calibration signal. As shown in FIG. 1, the waveform generator 10 receives peak-to-peak amplitude information at amplitude control 2, frequency information at frequency control 4, and DC-offset information at DC-offset control 6. The waveform generator 10 further receives shape information at shape control 8. The amplitude control 2 adjusts the magnitude of a generated signal. The frequency control 4 adjusts the frequency of the generated signal. The DC-offset control shifts the generated signal in a desired magnitude from a reference level. The shape control 8 adjusts the magnitude of the generated signal over time. In response to the amplitude control 2, the frequency control 4, the DC-offset control 6 and the shape control 8, the waveform generator 10 generates and provides a calibration signal at output 15 and renders a representation of the calibration signal on display 5. As indicated, the calibration signal should reflect accurately desired waveform characteristics for calibrating a test instrument.
The development of increasingly faster optical data transmission rates has generated a corresponding increase in the difficulty of accurately generating a calibration signal with desired characteristics. It is difficult and prohibitively expensive to generate a precise calibration signal with known DC and AC values or known relationships between DC and AC values because of distortion of the calibration signal at these higher operational frequencies and due to measurement error. Significant errors can be induced by measurement inaccuracies when generating such a prior art calibration signal. Additional errors can be introduced when measuring and processing the information in test equipment. Some of these additional errors are introduced through statistical processing of the acquired data and scaling of the received calibration signal to reduce inaccuracies introduced by the oscilloscope's analog-to-digital converter and other processing circuits.
Therefore, it would be desirable to provide an economical and reliable calibration solution that can be applied across a range of test instruments.